Today NASA announced the paradigm shifting discovery of flowing water on Mars. This extraterrestrial salty water bodes well for a water cycle on Mars, and potential hosting of Martian life. What mysteries lie on Mars, we may find out soon – but for the infinite mysteries that lie beyond – we have the Earth’s largest radio telescope, the Square Kilometre Array (SKA), manned by the Curtin Institute of Radio Astronomy.
The engineering challenges behind building the world’s biggest radio telescope are vast, but bring rewards beyond a better understanding of the universe.
Since its inception, the Curtin Institute of Radio Astronomy has established itself as an essential hub for astronomy research in Australia. Known as CIRA, the organisation brings together engineering and science expertise in one of Australia’s core research strengths: radio astronomy.
Through CIRA’s research node, Curtin is an equal partner in the International Centre for Radio Astronomy Research (ICRAR) with the University of Western Australia. Curtin also contributes staff to the Australian Research Council Centre of Excellence for All-sky Astrophysics. One of the core strengths of CIRA is the construction of next generation telescopes. These include work on one of the world’s biggest scientific endeavours and the SKA.
CIRA’s Co-Directors, Professors Steven Tingay and Peter Hall, were on the team who pitched Australia’s successful bid to host part of the SKA – a radio telescope that will stretch across Australia and Africa. The SKA’s two hosting nations were announced in May 2012 and the project forms the main focus of research at CIRA. And for good reason: the SKA-low – a low-frequency aperture array consisting of a quarter of a million individual antennas in its first phase – will be built in Western Australia at the Murchison Radio-astronomy Observatory (MRO), about 800 km north of Perth.
The near-flat terrain and lack of radio noise from electronics and broadcast media in this remote region allow for great sky access and ease of construction. At Phase 1, SKA-low will cover the project’s lowest-frequency band, from 50 MHz up to 350 MHz – with antennas covering approximately 2 km at the core, stretching out to 50 km along three spiral arms.
“Out of 10 organisations in a similar number of countries, CIRA is the largest single contributor to the low frequency array consortium,” says Hall, the Director responsible for engineering at CIRA.
Far from a traditional white dish radio telescope, which mechanically focuses beams, the SKA-low will be a huge array of electronic antennas with no moving parts. Its programmable signal processors will be able to focus on multiple fields of view and perform several different processes simultaneously. “You can point at as many directions as you want with full sensitivity – that’s the beauty of the electronic approach,” says Senior Research Fellow Dr Randall Wayth, an astronomer and signal processing specialist at CIRA.
One of the major scientific goals of SKA-low is to help illuminate the events of the early universe, particularly the stage of its formation known as the ‘epoch of reionisation’. Around 13 billion years ago, all matter in the early universe was ionised by radiation emitted from the earliest stars. The record of this reionisation carries with it telltale radio signatures that reveal how those early stars formed and turned into galaxies. Observing this directly for the first time will allow astronomers to unlock fundamental new physics.
“To see what’s going on there at the limits of where we can see in time and space, you have to have telescopes that are sensitive to wide-field, diffuse structures, and that are exquisitely calibrated. You have to be able to reject the foreground universe and local radio frequency interference,” says Hall. This sensitivity to diffuse structures will make SKA-low and its precursor, the Murchison Widefield Array (MWA), essential instruments in studying the epoch of reionisation.
The SKA-low will also be important in studying time domain astronomy, which consists of phenomena occurring over a vast range of timescales. One example is the field of pulsar study. Pulsars are incredibly dense rotating stars that, much like a lantern in a lighthouse, emit a beam of radiation at extremely regular intervals. This regularity makes pulsars useful tools for a variety of scientific applications, including accurate timekeeping.
By the time the radio signal from a distant pulsar travels across space and reaches Earth, it is dispersed. But with the right telescope, you can calibrate against this dispersion, and trace back the original regular signal.
“One of the great things you can do with a low frequency telescope such as the SKA-low is get a very good look at the pulsar signal,” says Hall. “As well as stand-alone SKA-low pulsar studies, the measurement of hour-to-hour dispersion changes can be fed to telescopes at higher frequencies, vastly improving their ability to do precision pulsar timing.”
“It’s a big advantage having the critical mass of people in this building to make things happen.”
It’s not just astronomy research that is benefiting from the construction of the SKA-low and its precursors (two precursor telescopes are in place at the MRO: the MWA and the Australian Square Kilometre Array Precursor telescope, ASKAP). In order to make the most out of the aperture array telescopes, some fundamental engineering challenges need to be solved. Challenges such as how to characterise the antennas to ensure that they meet design specifications, or how to design a photovoltaic system to power the SKA without producing too many unwanted emissions. Solving these problems requires both a deep understanding of the fundamental physics involved as well as knowledge of how to engineer solutions around those physics.
The projected construction timeframe for SKA-low is 2018–2023, but there is already infrastructure in place to begin testing its design and operation. Consisting of 2048 fixed dual-polarisation dipole antennas arranged in 128 ‘tiles’, the MWA boasts a wide field of view of several hundred square degrees at a resolution of arcminutes. It has provided insight into the challenges that will arise during the full deployment of SKA-low, not the least of which is managing the volume of data resulting from the measurements.
“The MWA already has a formidable data rate. We transmit 400 megabits per second down to Perth, and processing that is a substantial challenge,” says Wayth. The challenge is a necessary one, as the stream of data that comes from a fully operational SKA-low will be orders of magnitude larger.
“While doing groundbreaking science, the MWA is just manageable for us at the moment in terms of data rate. It teaches us what we have to do to handle the data.”
Continued CIRA developments at the MRO have included the construction of an independently commissioned prototype system, the Aperture Array Verification System 0.5 (AAVS0.5). The results from testing it in conjunction with the MWA surprised the engineers and scientists. “Engineers know that building even a tiny prototype teaches you a lot,” says Hall.
In their case, some carefully-matched cables turned out to be mismatched in their electrical delay lengths. Using the AAVS0.5, they have already been able to improve the MWA calibration. “We were able to feedback that engineering science into the MWA astronomy calibration model, and we now have a better model to calibrate and clean the images from the MWA,” says Hall.
Following the success of AAVS0.5, over the next two years CIRA will be leading the construction of the much larger AAVS1, designed to mimic a full SKA-low station.
Developing the SKA-low and its precursors is an huge effort, demanding the best in astrophysics, engineering and data processing. CIRA is uniquely positioned to accomplish this feat, with a large research staff, fully equipped engineering laboratory and access to the nearby Pawsey Supercomputing Centre for data processing. “CIRA has astronomers and engineers, as well as people who do both. We have all the skills to do these things in-house,” says Hall.
“It’s a big advantage having the critical mass of people in this building to make things happen,” says Wayth. “It’s a rare case where the sum of the parts really is greater than the whole.”
Opportunities for students and early-career researchers to engage in the project are already underway. Dozens of postgraduate research projects commencing in 2015 will involve the MWA, AAVS and ASKAP directly. Topics range from detecting the radio signature of fireballs to investigating the molecular chemistry of star formation. As well as producing novel scientific outcomes, these projects will feed valuable test data into the major scientific investigations slated for the SKA as it becomes operational.
A Supercomputer in the backyard
The scale of SKA, and the resultant flood of data, requires the rapid development of methods to process data. The Pawsey Supercomputing Centre – a purpose-built powerhouse named after pioneering Australian radio astronomer Dr Joe Pawsey and run by the Interactive Virtual Environments Centre (iVEC) – includes a supercomputer called Galaxy, dedicated to radio astronomy research. A key data challenge is finding ways in which the signal processing method can be split up and processed simultaneously, or ‘parallelised’, so that the full force of the supercomputing power can be used. The proximity of the signal processing experts at CIRA to iVEC means that researchers can continually prototype new ways of parallelising the data, with the goal being to achieve real-time analysis of data streaming in from the SKA.
– Phillip English